The Agent Skills Directory

[SAFE]: No malicious patterns detected. The skill consists entirely of Markdown documentation and examples designed to guide an AI agent in assisting developers with Java utility libraries.
[NO_CODE]: The skill does not include any scripts (.py, .js, .sh), executables, or dynamic configuration files. It acts purely as a reference for existing library code.
[DATA_EXPOSURE]: The skill includes a specific privacy and security section in SKILL.md that instructs the agent not to expose tokens, passwords, or private credentials in logs or exceptions.
[REMOTE_CODE_EXECUTION]: The documentation identifies and warns against potential RCE hazards in the ddd4j framework, specifically advising against using the typing-enabled REDIS_OBJECT_MAPPER for untrusted external JSON data.
[INDIRECT_PROMPT_INJECTION]: While the skill involves processing developer-provided JSON scenarios, it provides strict boundary instructions for the agent to distinguish between trusted and untrusted data sources to prevent injection-style attacks on the underlying application.

ddd4j-kit